The Smell of Molten Projects in the Morning

Ed Nisley's Blog: Shop notes, electronics, firmware, machinery, 3D printing, laser cuttery, and curiosities. Contents: 100% human thinking, 0% AI slop.

Author: Ed

Photo Backdrop Clamp Pad Embiggening

We got a photo backdrop stand to hold Mary’s show-n-tell quilts during her quilting club meetings, but the clamps intended to hold the backdrop from the top bar don’t work quite the way one might expect. These photos snagged from the listing shows their intended use:

Emart Photo Backdrop - clamp examples — Emart Photo Backdrop – clamp examples

The clamp closes on the top bar with the jaws about 15 mm apart, so you must wrap the backdrop around the bar, thereby concealing the top few inches of whatever you intended to show. This doesn’t matter for a preprinted generic backdrop or a green screen, but quilt borders have interesting detail.

The clamps need thicker jaws, which I promptly conjured from the vasty digital deep:

Spring Clamp Pads - PS preview — Spring Clamp Pads – PS preview

The original jaws fit neatly into those recesses, atop a snippet of carpet tape to prevent them from wandering off:

Spring Clamp pads - detail — Spring Clamp pads – detail

They’re thick enough to meet in the middle and make the clamp’s serrated round-ish opening fit around the bar:

Spring Clamp pads - compared — Spring Clamp pads – compared

With a quilt in place, the clamps slide freely along the bar:

Spring Clamp pads - fit test — Spring Clamp pads – fit test

That’s a recreation based on actual events, mostly because erecting the stand wasn’t going to happen for one photo.

To level set your expectations, the “Convenient Carry Bag” is more of a wrap than a bag, without enough fabric to completely surround its contents:

I put all the clamps / hooks / doodads in a quart Ziploc baggie, which seemed like a better idea than letting them rattle around loose inside the wrap. The flimsy pair (!) of hook-n-loop straps don’t reach across the gap and, even extended with a few inches of double-sided Velcro, lack enough mojo to hold it closed against all the contents.

It’ll suffice for our simple needs, but …

The OpenSCAD source code as a GitHub Gist:

	// Clamp pads for Emart photo backdrop clamps
	// Ed Nisley KE4ZNU Jan 2021

	/* [Hidden] */

	ThreadThick = 0.25;
	ThreadWidth = 0.40;

	HoleWindage = 0.2;

	Protrusion = 0.1; // make holes end cleanly

	inch = 25.4;

	function IntegerMultiple(Size,Unit) = Unit * ceil(Size / Unit);

	module PolyCyl(Dia,Height,ForceSides=0) { // based on nophead's polyholes
	Sides = (ForceSides != 0) ? ForceSides : (ceil(Dia) + 2);
	FixDia = Dia / cos(180/Sides);
	cylinder(d=(FixDia + HoleWindage),h=Height,$fn=Sides);
	}

	//———————-
	// Dimensions

	OEMpad = [24.0,16.0,3.0]; // original pad

	Pad = [35.0,25.0,8.0 + OEMpad.z]; // pad extension

	PadOffset = [0,-3.0,0];

	CornerRad = 3.0; // corner rounding

	Gap = 3.0;

	//———————-
	// Shape the pad

	module BigPad() {

	difference() {

	hull()
	for (i=[-1,1],j=[-1,1],k=[-1,1])
	translate([i(Pad.x/2 – CornerRad),j(Pad.y/2 – CornerRad),k*(Pad.z/2 – CornerRad) + Pad.z/2])
	sphere(r=CornerRad,$fn=6);

	translate(PadOffset + [0,0,Pad.z – (OEMpad.z + Protrusion)/2])
	cube(OEMpad + [HoleWindage,HoleWindage,Protrusion],center=true);

	}

	}

	//———————-
	// Build a pair

	translate([0,(Pad.y + Gap)/2,0])
	BigPad();

	translate([0,-(Pad.y + Gap)/2,0])
	rotate(180)
	BigPad();

view raw Spring Clamp Pads.scad hosted with ❤ by GitHub

2021-01-28

Removing Sentry FireSafe Door Trim

After turning the key and dialing the correct combination, the unlocking handle on the basement safe refused to move. Applying the dictum “If brute force isn’t working, you’re not using enough of it”, we eventually persuaded the handle downward and the door swung open. Before applying the dictum “If it doesn’t move and should, use WD-40 [*]”, I had to remove the trim cover from the interior side of the door to gain access to the lockwork.

Start by pressing the two latches inside the small circular holes in the hinge side of the trim cover:

Sentry FireSafe Door – first latches

Pull the trim over the slightly chamfered bolts, taking advantage of the fact the bottom pulls out more easily than the top, then pry it upward off the latch in the middle of the top:

Sentry FireSafe Door – top latch

The latches along the other side yield when you bend the trim sufficiently far from the door while applying the Designated Prydriver near the faint mold markings visible around the edge:

Sentry FireSafe Door – last latches

With the trim removed, you can see fancier safes in this line have two additional bolts engaging the top and bottom of the door frame. The rectangular block on the bottom of the hinge side might be for the batteries required by the spendy electronic lock version.

The problem turned out to be a lack of lubrication in the door-closed interlock that prevents you from relocking the bolts and wrecking the mechanism when you try to close the door with the bolts extended:

Sentry FireSafe Door – bolt interlock

A dab of silicone lube on all those sliding interfaces restored the interlock’s good humor.

The bolts had always rubbed just a bit on the trim cover, so I ovalized the offending side of the holes with an Xacto knife.

Slamming the trim firmly back on the door reset the latches, cycling the lockwork resynchronized the interlock, and the safe once again works just like it should.

[*] WD-40 is not the appropriate lube for the lockwork inside a safe, but the dictum aims you in the right direction.

2021-01-27
Tek Circuit Computer: Cursor Hairline

Given a machined cursor blank, clamp it into position:

Tek CC Cursor – cursor hairline fixture

You don’t want to clamp the cursor directly to the Sherline tooling plate, because the diamond drag bit would pass over two or three of those 10-32 screw holes which would, by the conservation of perversity, leave visible defects. In hindsight, I should have put a recess for an aluminum plate in there.

After a single pass at Z=-4.0 mm, add two strips of tape to protect the adjoining surface and scribble it with red lacquer crayon:

Tek CC Cursor – tape color fill

Peel the tape off:

Tek CC Cursor – tape removed

Then wipe off the residue using a soft cloth wetted with denatured alcohol:

Tek CC Cursor – red cursor detail

That looks much like the previous efforts. I’d like a more uniform trench, but I don’t know how to get there from here.

In any event, the hairline looks pretty good against laser-printed scales:

Tek CC Cursor – red cursor white laser decks – magnified

The new cursor is the lower one lying atop a laser-printed Pickett-style Circuit Computer:

Tek CC Cursor – red cursor yellow laser decks – overview

Looks good enough to eat, as the saying goes …

2021-01-26

Tek Circuit Computer: Cursor Milling Toolpath

Unlike the adhesive fixture, this setup requires a pause while milling the cursor outline to reclamp it from the front:

Tek CC Cursor Fixture - outline rear clamp — Tek CC Cursor Fixture – outline rear clamp

The trick is applying the front clamp before releasing the rear clamp:

Tek CC Cursor Fixture - outline both clamp — Tek CC Cursor Fixture – outline both clamp

Then continue the mission:

Tek CC Cursor Fixture - outline front clamp — Tek CC Cursor Fixture – outline front clamp

Because the tool path includes cutter compensation, GCMC adds entry and exit arcs to ensure a smooth transition:

Tek CC Cursor - Milling path — Tek CC Cursor – Milling path

The pix show a single cursor in the fixture while verifying the setup worked the way it should. Obviously, milling a stack of cursors eliminates a whole bunch of fiddling.

The tweaked MillCursor function from the mostly otherwise unchanged GCMC code:

    comment("Clamp on rear half of cursor!");

    local cp = {p0};                                             // enter at hub tangent point
    cp += varc_ccw([0mm,-2*p0.y,-],-hr,0,0.2mm,5deg) + p0;       // arc to tangent at hub bottom

    cp += {[p1.x,-p1.y,-]};                                      // lower tip entry point
    cp += varc_ccw([p2.x-p1.x,-(p2.y-p1.y),-],CursorTipRadius,0,0.2mm,5deg) + [p1.x,-p1.y,-];  // arc to tip exit at p2

    cp += varc_ccw([p1.x-p2.x,p1.y-p2.y,-],CursorTipRadius,0,0.2mm,5deg) + p2;  // arc to tip exit at p1

    goto([-,-,CursorSafeZ]);
    goto([0,0,-]);
    feedrate(MillSpeed);
    tracepath_comp(cp,CutterOD/2,TPC_OLDZ + TPC_RIGHT + TPC_ARCIN + TPC_ARCOUT);

    comment("Clamp on front half of cursor!");
    pause();                                      // wait for reclamping

    p1.z = MillZ;                                //  ... set milling depth
    cp = {p1};
    cp += {p0};
                                                 // exit at hub tangent
    tracepath_comp(cp,CutterOD/2,TPC_OLDZ + TPC_RIGHT + TPC_ARCIN + TPC_ARCOUT);

<<< snippage >>>

  goto([-,-,CursorSafeZ]);
  goto([0,0,-]);

Next, scribing a nice hairline with the new fixture.

2021-01-25

Tek Circuit Computer: 3D Printed Cursor Milling Fixture

The original Tektronix Circuit Computer cursor was probably die-cut from a larger sheet carrying pre-printed hairlines:

Tek CC - genuine - detail — Tek CC – genuine – detail

Machining a punch-and-die setup lies well beyond my capabilities, particularly given the ahem anticipated volume, so milling seems the only practical way to produce a few cursors.

Attaching a cursor blank to a fixture with sticky tape showed that the general idea worked reasonably well:

Tek CC - Cursor blank on fixture — Tek CC – Cursor blank on fixture

However, the tape didn’t have quite enough griptivity to hold the edges completely flat against milling forces (a downcut bit might have worked better) and I found myself chasing the cutter with a screwdriver to hold the cursor in place. Worse, the tape’s powerful attraction to swarf made it a single-use item.

Some tinkering showed a single screw in the (pre-drilled) pivot hole, without adhesive underneath, lacked enough oomph to keep the far end of the cursor in place, which meant I had to think about how to hold it down with real clamps.

Which, of course, meant conjuring a fixture from the vasty digital deep. The solid model includes the baseplate, two cutting templates, and a clamping fixture for engraving the cursor hairline:

Cursor Fixture - build layout — Cursor Fixture – build layout

The perimeter of the Clamp template on the far left is 0.5 mm inside the cursor perimeter. Needing only one Clamp, I could trace it on a piece of acrylic, bandsaw it pretty close, introduce it to Mr Belt Sander for final shaping, and finally drill the hole:

Tek CC Cursor Fixture - clamp drilling — Tek CC Cursor Fixture – clamp drilling

The Rough template is 1.0 mm outside the cursor perimeter, so I can trace those outlines on a PET sheet:

Tek CC Cursor Fixture - Rough template layout — Tek CC Cursor Fixture – Rough template layout

Then cut the patterns with a scissors, stack ’em up, and tape the edges to keep them aligned:

TekCC Cursor Fixture - Rough template — TekCC Cursor Fixture – Rough template

Align the stack by feel, apply the Clamp to hold them in place, and secure the stack with a Sherline clamp:

The alert reader will note it’s no longer possible to machine the entire perimeter in one pass; more on that in a while.

The baseplate pretty much fills the entire Sherline tooling plate. It sports several alignment pips at known offsets from the origin at the center of the pivot hole:

Tek CC Cursor Fixture - touch-off point — Tek CC Cursor Fixture – touch-off point

Dropping the laser alignment dot into a convenient pip, then touching off X and Y to the known offset sets the origin without measuring anything. Four screws in the corners align the plate well enough to not worry about angular tweakage.

The OpenSCAD source code as a GitHub Gist:

	// Machining fixtures for Tek Circuit Computer cursor
	// Ed Nisley KE4ZNU Jan 2021

	Layout = "Show"; // [Show, Build, Cursor, Clamp, Rough, Engrave]

	/* [Hidden] */

	ThreadThick = 0.25;
	ThreadWidth = 0.40;

	HoleWindage = 0.2;

	Protrusion = 0.1; // make holes end cleanly

	inch = 25.4;

	function IntegerMultiple(Size,Unit) = Unit * ceil(Size / Unit);

	module PolyCyl(Dia,Height,ForceSides=0) { // based on nophead's polyholes
	Sides = (ForceSides != 0) ? ForceSides : (ceil(Dia) + 2);
	FixDia = Dia / cos(180/Sides);
	cylinder(d=(FixDia + HoleWindage),h=Height,$fn=Sides);
	}

	//———————-
	// Dimensions

	CursorHubOD = 1.0*inch; // original Tek CC was hard inch!
	CursorTipWidth = (9.0/16.0)*inch;
	CursorTipRadius = (1.0/16.0)*inch;

	CursorThick = 0.5; // plastic sheet thickness

	CutterOD = 3.175; // milling cutter dia
	CutterDepth = 2.0; // … depth of cut
	CutterLip = 0.5; // … clearance under edge

	ScribeOD = 3.0; // diamond scribe shank

	StudOC = [1.16inch,1.16inch]; // Sherline tooling plate grid
	StudClear = 5.0; // … screw clearance
	StudWasher = 11.0; // … washer OD

	CursorOffset = [-2*StudOC.x,0,0]; // hub center relative to fixture center

	// must have even multiples of stud spacing to put studs along centerlines
	BasePlateStuds = [6StudOC.x,2StudOC.y]; // fixture screws
	echo(str("Stud spacing: ",StudOC));

	CornerRad = 10.0; // corner radius

	BasePlate = [2StudWasher + BasePlateStuds.x,2StudWasher + BasePlateStuds.y,5.0];
	echo(str("Base Plate: ",BasePlate));

	EngravePlate = [5StudOC.x,1.5StudOC.y,BasePlate.z];
	echo(str("Engrave Plate: ",EngravePlate));

	TemplateThick = 6*ThreadThick;
	LegendThick = 2*ThreadThick;

	Gap = 3.0;

	//———————-
	// Import SVG of cursor outline
	// Requires our hub OD to match reality
	// Hub center at origin

	module CursorSVG(t=CursorThick,od=0) {

	hr = CursorHubOD/2;

	translate([-hr,-hr,0])
	linear_extrude(height=t,convexity=3)
	offset(r=od/2)
	import(file="/mnt/bulkdata/Project Files/Tektronix Circuit Computer/Firmware/TekCC-Cursor-Mark.svg",center=false);

	}

	//———————-
	// Milling fixture for cursor blanks

	module Fixture() {

	difference() {
	hull() // basic plate shape
	for (i=[-1,1], j=[-1,1])
	translate([i(BasePlate.x/2 – CornerRad),j(BasePlate.y/2 – CornerRad),0])
	cylinder(r=CornerRad,h=BasePlate.z,$fn=24);

	translate(CursorOffset + [0,0,BasePlate.z – CutterDepth])
	difference() {
	CursorSVG(CutterDepth + Protrusion,1.5*CutterOD);
	CursorSVG(CutterDepth + Protrusion,-CutterLip);
	}

	translate(CursorOffset + [0,0,BasePlate.z – 2*ThreadThick]) { // alignment pips
	for (x=[-20.0,130.0], y=[-30.0,0.0,30.0])
	translate([x,y,0])
	cylinder(d=4*ThreadWidth,h=1,$fn=6);

	for (x=[-30.0,130.0,150.0])
	translate([x,0,0])
	cylinder(d=4*ThreadWidth,h=1,$fn=6);
	}

	for (i=[-1,1], j=[-1,1]) // mounting stud holes
	translate([iBasePlateStuds.x/2,jBasePlateStuds.y/2,-Protrusion])
	rotate(180/6)
	PolyCyl(StudClear,BasePlate.z + 2*Protrusion,6);

	translate(CursorOffset + [0,0,-Protrusion]) // hub clamp hole
	rotate(180/6)
	PolyCyl(StudClear,BasePlate.z + 2*Protrusion,6);

	translate([2*StudOC.x,0,-Protrusion]) // tip clamp hole
	rotate(180/6)
	PolyCyl(StudClear,BasePlate.z + 2*Protrusion,6);

	for (i=[-2:2], j=[-1,1]) // side clamp holes
	translate([iStudOC.x,jStudOC.y,-Protrusion])
	rotate(180/6)
	PolyCyl(StudClear,BasePlate.z + 2*Protrusion,6);

	}

	}

	//———————-
	// Show-n-Tell cursor

	module Cursor() {

	difference() {
	CursorSVG(CursorThick,0.0);
	translate([0,0,-Protrusion])
	rotate(180/6)
	PolyCyl(StudClear,TemplateThick + 2*Protrusion,6);
	}
	}


	//———————-
	// Template for rough-cutting blanks

	module Rough() {

	bb = [40,12,LegendThick];

	difference() {
	CursorSVG(TemplateThick,1.0);
	translate([0,0,-Protrusion])
	rotate(180/6)
	PolyCyl(StudClear,TemplateThick + 2*Protrusion,6);

	difference() {
	translate([bb.x/2 + CursorHubOD/2,0,TemplateThick – bb.z/2 + Protrusion])
	cube(bb + [0,0,Protrusion],center=true);
	translate([bb.x/2 + CursorHubOD/2,0,TemplateThick – bb.z])
	linear_extrude(height=bb.z,convexity=10)
	text(text="Rough",size=7,spacing=1.00,font="DejaVu Sans:style:Bold",halign="center",valign="center");
	}
	}
	}

	//———————-
	// Template for aluminium clamping plate

	module Clamp() {

	bb = [40,12,LegendThick];

	difference() {
	CursorSVG(TemplateThick,-1.0);
	translate([0,0,-Protrusion])
	rotate(180/6)
	PolyCyl(StudClear,TemplateThick + 2*Protrusion,6);

	difference() {
	translate([bb.x/2 + CursorHubOD/2,0,TemplateThick – bb.z/2 + Protrusion])
	cube(bb + [0,0,Protrusion],center=true);
	translate([bb.x/2 + CursorHubOD/2,0,TemplateThick – bb.z])
	linear_extrude(height=bb.z,convexity=10)
	text(text="Clamp",size=7,spacing=1.00,font="DejaVu Sans:style:Bold",halign="center",valign="center");
	}
	}
	}

	//———————-
	// Engraving clamp

	module Engrave() {

	difference() {

	hull() // clamp outline
	for (i=[-1,1], j=[-1,1])
	translate([i(EngravePlate.x/2 – CornerRad),j(EngravePlate.y/2 – CornerRad),0])
	cylinder(r=CornerRad,h=EngravePlate.z,$fn=24);

	translate(CursorOffset + [0,0,-Protrusion])
	CursorSVG(CursorThick + Protrusion,0.5); // pocket for blank cursor

	translate(CursorOffset + [0,0,-Protrusion])
	rotate(180/6)
	PolyCyl(StudClear,EngravePlate.z + 2*Protrusion,6);

	translate([2*StudOC.x,0,-Protrusion])
	rotate(180/6)
	PolyCyl(StudClear,EngravePlate.z + 2*Protrusion,6);

	hull() {
	for (i=[-1,1])
	translate([i1.5StudOC.x,0,-Protrusion])
	PolyCyl(2ScribeOD,EngravePlate.z + 2Protrusion,8);
	}
	}

	}


	//———————-
	// Build it

	if (Layout == "Cursor") {
	Cursor();
	}

	if (Layout == "Clamp") {
	Clamp();
	}

	if (Layout == "Rough") {
	Rough();
	}

	if (Layout == "Engrave") {
	Engrave();
	}

	if (Layout == "Show") {
	Fixture();
	color("Green",0.3)
	translate(CursorOffset + [0,0,BasePlate.z + Protrusion])
	Cursor();

	color("Orange")
	translate(CursorOffset + [0,0,BasePlate.z + 10])
	Rough();


	color("Brown")
	translate(CursorOffset + [0,0,BasePlate.z + 20])
	Clamp();

	color("Gold")
	translate(0*CursorOffset + [0,0,BasePlate.z + 40])
	Engrave();

	}

	if (Layout == "Build"){
	rotate(90) {
	Fixture();
	translate([0,-((BasePlate.y + EngravePlate.y)/2 + Gap),EngravePlate.z])
	rotate([180,0,0])
	Engrave();
	translate(CursorOffset + [0,(BasePlate.y + CursorHubOD)/2 + Gap,0])
	Rough();
	translate(CursorOffset + [0,(BasePlate.y + 3CursorHubOD)/2 + 2Gap,0])
	Clamp();
	}
	}

view raw Cursor Fixture.scad hosted with ❤ by GitHub

The original doodle with some notions and dimensions that didn’t survive contact with reality:

I have no idea why the Sherline tooling plate has a 10-32 screw grid on 1.16 inch = 29.46 mm centers, but there they are.

2021-01-22

Astable Multivibrator: Red RGB Piranha

A red LED has a sufficiently low forward voltage to run with a MOSFET astable multivibrator and a pair of run-down AA alkaline cells:

Astable AA Alkaline – red

The red LED is actually part of an RGB Piranha, just to see how it compares to an as-yet-unbuilt version with a single red LED in the same package.

The LED drops 1.9 V of the 2.75 V from the mostly used-up AA cells:

Astable Piranha Red – 2.75 alkaline – V LED

The original 33 Ω ballast resistor showed a peak current of 11 mA in a 30 ms pulse:

Astable Piranha Red – 2.75 alkaline – V 33 ohm

Replacing it with a 12 Ω resistor boosts the current all the way to 12 mA:

Astable Piranha Red – 2.75 alkaline – V 12 ohm

The 2N7000 gate sees a just bit more than 2 V, barely enough to get the poor thing conducting, which makes the ballast resistor mostly decorative. The MOSFET datasheet puts its 1 mA threshold somewhere between 0.8 and 3 V, so it could be worse.

Keep in mind the DSO150’s 1 MΩ input impedance sat in parallel with the 1 MΩ gate pulldown resistor forming the RC differentiator when I measured the gate voltage; I’ll leave the simulation as an exercise for the interested reader. The blinks were noticeably dimmer and perhaps a bit shorter, although eyeballometric calibration is notoriously hard.

The slightly revised schematic-layout doodle stacks the transistors along the negative bus bar:

Astable wiring layout – stacked 2N7000

Flipping the bottom transistor over to snuggle the two timing caps next to each other would eliminate the long jumper wire and probably look better.

2021-01-21
Homage Tektronix Circuit Computer: Laser Printed Scales

Given the proper command-line options, GCMC can produce an SVG image and, after some Bash fiddling and a bank shot off Inkscape, the same GCMC program I’ve been using to plot Homage Tektronix Circuit Computer decks can produce laser-printed decks:

Tek CC – laser – detail

Pen-plotting on yellow Astrobrights paper showed how much ink bleeds on slightly porous paper, but laser-printing the same paper produces crisp lines:

Tek CC – laser – yellow detail

Laser printing definitely feels like cheating, but, for comparison, here’s a Genuine Tektronix Circuit Computer:

Tek CC – genuine – detail

Plotting the decks on hard mode was definitely a learning experience!

Obviously, my cursor engraving hand remains weak.

2021-01-20